Why this change was accepted

Hi

Felix Fietkau has turned of the diagnostics for
STA txq pending airtime underflow:
https://www.spinics.net/lists/linux-wireless/msg201168.html

and he works/worked for/with MediaTek, example with its copyrights is below:
https://patchwork.kernel.org/patch/11496331/

but the "STA txq pending airtime underflow" still wasn't fixed and when it happens the only solution is to reboot the router... (otherwise it scrambles the traffic, decreases the speed and so on)

Can somebody explain what is going on?

I think you already wrote what is happening on a technical side.
Or were you asking so very kindly why the current development HEAD is not fully functional for you requirements after you downloaded it for free of the interwebz ? I am somewhat confused by your rather explosive post. Were you asking for help or just venting anger ? certainly not a very productive/invitational beginning for trying to resolve an open source issue.

3 Likes

I think you already wrote what is happening on a technical side.

Or were you asking so very kindly why the current development HEAD is not fully functional for you requirements after you downloaded it for free of the interwebz?
No i'm not

I am somewhat confused by your rather explosive post.
I don't think it is more explosive then than the discussed patch.

Were you asking for help or just venting anger?
I'm asking why the diagnostics message used to mark the issue in mac80211 had been removed without the fix of the issue?

certainly not a very productive/invitational beginning for trying to resolve an open >>source issue.
It was observed a long time ago. And instead of fixing it the MediaTek developer just removed the warning (if you checked the links)

Ok, so now we are responsible for $MTK dev doing something upstream ? makes no sense to me. regardless of that, be so kind to send a friendly/productive mail to the involved devs. You will find that kindness gets you way further than flaming and naming. we are a big family after all, its about the code, not the person.
best regards
John

1 Like

Sorry. Did you get the content of mentioned patch?

did you get the implied advice of my reply ?

1 Like

Copyright (C) 2014 Felix Fietkau <nbd@openwrt.org>
So, You're here because of a 6 year old tag line.

Did you try emailing that address?

Or, did you search the forum for @nbd?

1 Like

I removed the warning because it was bogus. You’ve been seeing stability issues at the same time as these warnings appeared, because both were triggered by disassociation and reassociation. That does not imply that the condition that triggered the warning was related to the problem in any way.
In fact, I tracked down the stability issues to a firmware bug, resolved by using a different API for adding/removing stations. The fix is in OpenWrt master, please test it.

1 Like